Microsoft on Monday announced the launch of Project Perception, an agentic security system designed to help defenders combat a rapidly evolving threat of AI-based attacks by malicious actors.
The company also introduced MAI-Cyber-1-Flash, an AI model that helps users identify and remediate vulnerabilities through an integration with the company’s MDASH code-scanning harness.
The launch is part of a major industry push to implement AI-based security that can protect corporate IT and critical infrastructure systems from sophisticated attacks. In the past year, threat actors have demonstrated the ability to speed and scale attacks in a way that traditional security technologies cannot match.
“So, if you think about it, we’re now in a world where we’ve moved from AI-assisted attacks to AI-operated attacks to now autonomous AI attacking customers,” Hayete Gallot, executive vice president of Microsoft Security, said at a San Francisco launch event Monday. “The physics of cyber have fundamentally changed.”
Project Perception combines a group of red team, blue team and green team agents to help simulate, detect and remediate potential vulnerabilities and other risks inside an organization’s environment before bad actors can discover the same paths.
The accelerated threat of agentic AI makes the launch even more urgent, because these autonomous systems have since developed the ability to find and create attack paths that can operate with almost no human interaction.
Microsoft is in a unique position. It collects trillions of signals that can be used to detect potential threat activity. In addition, the company said its system provides cost savings over rival models currently available.
“Launching their own model, based on their own data and expertise, ensures the model is best suited to reason over Microsoft data and best aligns to its products,” Allie Mellen, principal analyst at Forrester, told Cybersecurity Dive.
The development of frontier AI models are enabling security teams to identify vulnerabilities at a volume and speed never before seen. At the same time, criminal and state-backed hackers have rapidly adapted their ability to identify and weaponize vulnerabilities faster than ever.
In May, Google Threat Intelligence Group reported that hackers used AI to develop a working zero-day exploit in an effort to launch a mass event. Researchers were able to notify the relevant developer and prevent a larger attack, but the incident served as a wake-up call that mass AI exploitation would soon become a legitimate risk in the near future.
